Part 9
Where to start
A compact decision aid by role and situation.
| You are… | Your situation | Start with |
|---|---|---|
| A clinician or ward leader | Recurring safety or quality problem | PDSA at tiny scale, then just culture and AAR |
| A clinical director | A pathway crossing departments is failing | Systems thinking for the flow, then future search |
| A nurse or junior doctor | You cannot raise concerns safely | Graded assertiveness now; psychological safety as the longer campaign |
| A health administrator | Staff survey results are poor | Survey feedback cascade, team by team; not another corporate program |
| An engineer | Delivery is slow and nobody knows why | DORA to find the constraint, then force field analysis |
| An engineering manager | The team is capable and unhappy | Job characteristics model and SDT, before any engagement initiative |
| A tech lead | Boundaries between teams cause constant coordination | Conway's law and team topologies |
| A CTO / CIO | A transformation has stalled | Beckhard-Harris to find the missing term, then McKinsey influence as an audit |
| A CEO | You suspect a culture problem | Rule out structure, incentives, and workload first via Burke-Litwin; then Schein |
| A CHRO | You are asked for a values relaunch | How to actually change culture — start with promotion criteria, not posters |
| Any executive | Two directors are in open conflict | Interests not positions; do not run team building |
| Any leader | You are about to reorganize | Galbraith star — design processes, rewards, and people moves before announcing structure |
| Any practitioner | You have been handed a mess with no brief | Entry and contracting, properly, before anything else |
A universal first ninety days, if you have no better plan:
- Weeks 1–2: contract explicitly. Read the last year of incident reviews, survey results, and board papers. Ask new joiners what surprised them.
- Weeks 3–5: twelve to twenty interviews plus direct observation of the actual work. Count things.
- Week 6: pick a diagnostic frame and populate it. Name whether the problem is transformational or transactional.
- Weeks 7–8: feed back to participants first, in their own groups. Let them interpret.
- Weeks 9–10: agree one intervention with a named owner, a date, a prediction, and a balancing measure.
- Weeks 11–13: run it small. Study the result against the prediction. Decide: adopt, adapt, or abandon.